Metabolic reprogramming is considered to be a hallmark of malignancy and source of novel therapeutic targets. However, the vast majority of knowledge so far established for tumor cell metabolism is derived from studies in cultured cell lines rather than intact tumors. Our work over the previous five years of this Award demonstrated that a) tumor cells contain a diversity of metabolic programs to support survival and proliferation, with many distinct configurations of the tricarboxylic acid (TCA) cycle regulate by combinations of cell-intrinsic and extrinsic factors; b) intra-operative infusions of 13C-glucos and other isotope-labeled fuels can be used to probe the metabolism of intact tumors in mice and humans; c) important metabolic differences exist between tumor cells in culture and in vivo, indicating a need to improve techniques for in vivo analysis, particularly in humans; d) human lung tumors support their bioenergetics by oxidizing glucose and a variety of other substrates in vivo; and e) pre-operative MRI, FDG-PET and other imaging techniques can be used to predict informative aspects of tumor metabolism and identify areas of metabolic heterogeneity within individual human lung tumors. We are now poised to use a unique and highly innovative combination of approaches to test hypotheses about the metabolism of intact tumors in humans and mice, focusing on the drivers of metabolic heterogeneity in vivo. We propose three Specific Aims to address these issues in non-small cell lung cancer (NSCLC), the most common cause of cancer-related deaths worldwide. In Aim 1, we will implement a novel technique to derive positional assignment of 13C by mass spectrometry. This will enable us to maximize the information content derived from very small samples, potentially as little as 1% the size of fragments currently required by standard approaches. This will greatly enhance our ability to map regional metabolic heterogeneity within individual tumors in vivo. In Aim 2, we will use pre-surgical imaging, 13C infusions and metabolomics to examine the metabolism of glucose and other fuels in intact tumors. We will determine how tissue perfusion alters nutrient preferences in vivo and test the hypothesis that enhanced glucose oxidation in the TCA cycle is a specific hallmark of proliferating tumor cells. This aim will use xenografts derived from NSCLC cell lines, patient-derived NSCLC xenografts, and a clinical study featuring intra-operative 13C infusions in human NSCLC patients. In Aim 3, we will test the hypothesis that lactate, an abundant circulating fuel long considered a waste product of tumor metabolism, is also used as a fuel source for the TCA cycle in a subset of well-perfused NSCLC tumors. We will use infusions of 13C-labeled lactate to identify and localize pathways of lactate utilization within intact tumors i humans and mice. Altogether, these Aims will generate a unique view of NSCLC metabolism with an unprecedented level of detail, biological accuracy and relevance to human disease. They have the potential to establish new paradigms in metabolic regulation and heterogeneity in cancer and in predicting which tumors will respond to metabolic therapies.